Gut 1996; 39: 214-219

Influence of peppermint oil on absorptive andsecretory processes in rat small intestine

A Beesley, J Hardcastle, P T Hardcastle, C J Taylor

AbstractBackground-Peppermint oil is used torelieve the symptoms of irritable bowelsyndrome, relaxing intestinal smoothmuscle by reducing the availability ofcalcium, but its effects on intestinaltransport are unknown.Aims-To determine the effect ofpepper-mint oil on intestinal transport processes.Methods-The influence ofpeppermint oilon intestinal transport was investigated inrat jejunum using both intestinal sheetsmounted in Ussing chambers and brushborder membrane vesicles.Results-Mucosal peppermint oil (1 and5 mg/ml) had no significant effect onbasal short circuit current, but inhibitedthe increase associated with sodiumdependent glucose absorption. Theincreased short circuit current induced byserosal acetylcholine, a reflection ofcalcium mediated electrogenic chloridesecretion, was unaffected by mucosalpeppermint oil (5 mg/ml). In contrast,serosal peppermint oil (1 mg/ml) inhibitedthe response to acetylcholine withoutreducing the effect of mucosal glucose.In brush border membrane vesiclesactive glucose uptake was inhibited byextravesicular peppermint oil at con-centrations of05 and 1 mg/ml.Conclusions-Peppermint oil in the intes-tinal lumen inhibits enterocyte glucoseuptake via a direct action at the brushborder membrane. Inhibition of secretionby serosal peppermint oil is consistent witha reduced availability ofcalcium.(Gut 1996; 39: 214-219)

Keywords: peppennint oil, small intestine, glucoseabsorption, chloride secretion, brush bordermembrane vesicles, rat.

(4.40/o).3 It is given orally, usually in entericcoated capsules that prevent its release in thestomach.8 The free oil will therefore first comeinto contact with the epithelial cells that linethe intestinal lumen. These enterocytes areresponsible for the absorption of nutrients,ions, and water from the intestinal lumen andthe possibility that peppermint oil might in-fluence these processes has not been previouslyinvestigated.

The enterocytes also possess the capacity forsecretion, a process that entails the stimulationof electrogenic chloride secretion, together withthe inhibition of electroneutral sodium chlorideabsorption.9 Intestinal secretion is regulated bya variety of neural,10 humoral," and immunefactors,'2 many of which change the level ofcytoplasmic calcium by increasing its entry fromthe extracellular compartment.'3 14 As pepper-mint oil has been shown to block calciummediated events in intestinal smooth muscle,7 itmay also inhibit intestinal secretion induced byagents that act in this way.

This study was designed to investigate thepossibility that peppermint oil can influencetransport across the small intestine by examin-ing its effects on both absorptive and secretoryprocesses. A preliminary report of the data hasbeen presented to the British Society ofGastroenterology. 15

Methods

AnimalsExperiments were carried out on male Wistarrats, 230-250 g, obtained from the SheffieldField Laboratories and permitted free access tofood and water. They were anaesthetised withsodium pentobarbitone (Sagatal, 60 mg/kgintraperitoneally).

Departments ofBiomedical ScienceA BeesleyJ HardcastleP T Hardcastle

and PaediatricsC J Taylor

Sheffield University

Correspondence to:Dr J Hardcastle, Departmentof Biomedical Science, TheUniversity, Western Bank,Sheffield S10 2TN.

Accepted for publication11 March 1996

Peppermint oil is a naturally occurring carmi-native, which is used to relieve the abdominalcolic and distension associated with irritablebowel syndrome.' It acts by relaxing intestinalsmooth muscle, an effect that has beenobserved both in vitro2-4 and in vivo.5Relaxation of smooth muscle by peppermintoil has been attributed to its ability to reducecalcium availability6 and a patch clamp studyhas shown that it acts on potential dependentcalcium channels in the muscle membrane,reducing the peak current amplitude andincreasing the rate of current decay.7

Peppermint oil consists of a number of con-stituents including menthol (44%), menthone(33%), cineole (12%), and menthyl acetate

Measurement of transintestinal electrical activityThe potential difference (PD), short circuitcurrent (SCC), and tissue resistance weremeasured in paired sheets taken from adjacentsegments of the mid-region of the smallintestine from which the outer muscle layersand myenteric plexus had been removed. Eachsheet was mounted in an Ussing chamber withan aperture of 1.925 cm2 and incubated at37°C in Kreb's bicarbonate saline gassed with95% 02- 5% CO2. The serosal fluid contained10 mM glucose and the mucosal fluid 10 mMmannitol, each having a volume of 5 ml. ThePD was measured using salt bridge electrodesconnected via calomel half cells to a differentialinput electrometer with output to a two

214

on May 21, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.39.2.214 on 1 A

ugust 1996. Dow

nloaded from

Influence ofpeppermint oil on absorptive and secretory processes in rat small intestine

channel chart recorder (Linseis L65 12).Current was applied across the tissue viaconductive plastic electrodes and the SCCmeasured as described by Field et al. 16

After mounting sheets were allowed tostabilise for 10 minutes after which readingsof PD and SCC were taken at one minuteintervals. After five minute basal readingspeppermint oil was added to either themucosal or serosal solution of the test sheet,with the control sheet receiving an equivalentvolume of the vehicle (1% ethanol v/v). Aftera further 10 minutes glucose, glycine oracetylcholine was added to both sheets.Glucose or glycine was added sequentially tothe mucosal solution to give concentrations of2.5, 5, 10, 15, 20, 25, 30, 35, and 40 mM. Toassess the effect of the increased osmolarity ofthe mucosal solution resulting from the addi-tion of nutrient, control experiments werecarried out in which nutrient was replacedwith equimolar mannitol. The rise in SCCassociated with active nutrient transport wastaken as the difference between the increaseseen with glucose and the decrease obtainedwith the same concentration of mannitol.Acetylcholine was added to the serosalsolution to give a concentration of 10-3M.

Measurement ofglucose uptake by brush bordermembrane vesiclesVesicles were prepared from the small intestineusing a method based on the magnesiumprecipitation technique.'7 Mucosal scrapes wereobtained from the jejuna (the first 25 cm fromthe ligament of Treitz) of two rats and stored at-80°C. To prepare vesicles the scrapes wereeach thawed in 18 ml buffer A(50 mM mannitol, 2 mM Na HEPES 0.02%NaN3, pH 7-4 at 4°C) and homogenised with aChemap El-Vibromixer (glass P1 plate, 3 cmdiameter) for two minutes at full speed(60 Hz). The homogenates from the two jejunawere then combined and aliquots taken forenzyme assays before adding 1M MgCl2 to afinal concentration of 10mM and stirring gentlyon ice for 20 minutes. Subsequent steps were allcarried out at 4°C. The mixture was centrifugedat 3000 g for 10 minutes and the supernatantrecentrifuged at 27 000 g for 30 minutes. Thebrush border membrane pellet was resuspendedin 20 ml buffer B (300 mM mannitol, 20 mMNa HEPES, 0.1 mM MgSO4, 0.02% NaN3,pH 7A4 at 20°C) by passing several timesthrough a 21 gauge needle. The suspension wasthen centrifuged for 15 minutes at 6000 g andthe supernatant recentrifuged at 27 000 g for afurther 30 minutes. The pellet consisting ofpurified brush border membrane vesicles wasfinally resuspended in buffer B using a 21 gaugeneedle to give a protein concentration of 10-18mg/ml.

Glucose uptake into freshly preparedvesicles was measured at 20°C using a filtrationstop technique. Three p1l vesicle suspensionwas added to 60 p1 incubation medium to givean extravesicular composition of 100 mMNaSCN, 100 ,uM D-[3H]-glucose (4 GBqmmol- 1), 120 mM mannitol, 20 mM Na

HEPES, 0.1 mM MgSO4, 0.02% NaN3, pH7.4 at 20°C. Passive uptake of glucose wasdetermined in the presence of 0.25 mM phlor-rhizin and subtracted from total uptake toprovide a measure of active glucose uptake.Peppermint oil or its vehicle (1% ethanol v/v)was added to the incubation medium to givethe concentrations indicated. When present,Triton X-100 was diluted with water andadded to the incubation medium to give a finalconcentration of 0.04%. Each experimentalvalue was taken as the mean of duplicatemeasurements made on a single sample.

Uptake of glucose was terminated at set timepoints by the addition of 2.5 ml ice-cold stopsolution (165 mM NaCl, 20 mM Na HEPES,0.25 mM phlorrhizin, 0.02% NaN3, pH 7-4 at4°C) and rapid filtration through a 0.45 urmnitrocellulose membrane filter (Whatman Ltd,Maidstone) under vacuum. The filters andincubation tubes were washed with a further5 ml stop solution. The filters were dissolved in5 ml Emulsifier Safe scintillation fluid for sub-sequent counting in a liquid scintillationanalyser (Packard Tri-carb 1600 TR), with allcounts being corrected for quenching. Non-specific binding of glucose to filter papers wasestimated by filtering incubation medium withno added vesicles. Glucose uptake is expressedas pmol/mg protein.

Protein was assayed by its capacity to bind toCoomassie blue using the Biorad assay tech-nique, with bovine serum albumin as standard.

Measurement of non-specific glucose binding tobrush border membrane vesiclesIt is possible that artificially high glucose uptakerates may be seen as a result of non-specificbinding to the external face of the brush bordermembrane vesicles. As the size of the intra-vesicular space is inversely proportional to theexternal osmolality, uptake of glucose at an infi-nitely high extravesicular osmolality must becaused by such non-specific binding.17 Theeffect of external medium osmolality uponapparent glucose uptake at equilibrium wastherefore assessed by incubating vesicles withextravesicular solutions containing differingamounts of D-cellobiose (0 to 600 mM) to alterosmolality. Each reaction was allowed to pro-ceed for 30 minutes to ensure that it hadreached equilibrium before halting it with theaddition of a stop solution that was isoosmoticwith the incubation medium. Vesicles were pre-pared with the following internal composition:100 mM mannitol or cellobiose, 20 mM NaHEPES, 01 mM MgSO4, 0-02% NaN3 (pH7-4 at 20°C). The choice of mannitol or cello-biose as the impermeant solute in the internalsolution did not significantly change the results.The final composition ofthe extravesicular solu-tion was 50 mM NaSCN, 20 mM Na HEPES,100 uM D-[3H]-glucose (4 GBq mmol- 1), 0.1mM MgSO4, 0.02% NaN3 (pH 7-4 at 20°C)plus varying concentrations of D-cellobiose andin all cases the osmolality was measured. Stopsolutions consisted of 50 mM NaCl, 20 mM NaHEPES, 0-25 mM phlorrhizim, 0.02% NaN3plus 0-600 mM mannitol (pH 7.4 at 4°C).

215

on May 21, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.39.2.214 on 1 A

ugust 1996. Dow

nloaded from

Beesley, Hardcastle, Hardcastle, Taylor

cm

= 100UU

cl)

0 10 20 30 40Concentration (mM)

Figure 1: Effect ofpeppermint oil on the changes in SCC induced by the serial addition ofglucose (solid symbols) or mannitol (open symbols) to the solution bathing the mucosalsurface of stripped sheets of rat small intestine. Peppermint oil was added to giveconcentrations of 0.1 (3 n=5, 0 n=6), 0.5 (A n=6, A n=6), 1 (V n=8, V n=6) or 5(O n= 10, 0 n=8) mg/ml, with control sheets (0 n=29, 0 n=26) receiving anequivalent volume (1 % v/v) of the ethanol vehicle. Each point represents the mean (SEM)of the number ofobservations indicated.

Assay of marker enzymesThe activities of brush border, basolateral, andmitochondrial marker enzymes were deter-mined in the initial homogenate and in thebrush border membrane vesicle suspension toevaluate the degree of purification of the finalpreparation. Alkaline phosphatase activity wasmeasured using a commercial kit (Sigma 104)with p-nitrophenyl phosphate as substrate.Sucrase activity was determined using amethod based on that of Dahlqvist,18 whilebasolateral Na+/K+-ATPase activity wasassayed as described by Hardcastle et al.19Succinate dehydrogenase activity wasmeasured as described by Pennington.20 Allenzyme activities were calculated per mgprotein.

ChemicalsPeppermint oil BP was obtained from Thorntonand Ross, Huddersfield; D-glucose and glycinefrom BDH Chemicals, Poole; phlorrhizindihydrate from Phase Separations, Queensferry;adenosine 5'-triphosphate (disodium salt), D-cellobiose, and acetylcholine chloride fromSigma Chemical, Poole. D-[3H]-glucose waspurchased from Amersham International,Amersham. The scintillation fluid used wasEmulsifier Safe from Canberra Packard,

Effect ofpeppermint oil on the SCCmax for glucose and glycine in stripped sheets of ratmid-intestine. The rise in SCC induced by glucose or glycine was correctedfor the osmoticcomponent by subtracting the change in SCC induced by equimolar mannitol. Peppermintoil was added to either the mucosal or serosal solution of test sheets as indicated, withcontrol sheets receiving an equivalent volume of vehicle (1% ethanol v/v). The SCCmaxvalues were determined using the Eadie-Hofstee plot and each value represents the mean(SEM) of the number of observations (n) indicated, with a paired t test being used toassess the significance ofpeppermint oil action. *p<O.OO1 (control v test)

SCCmax (pA/cm2)Peppermint oil(mg/mi) Application Number Control Test

Glucose 0.1 Mucosal 5 428 (22) 400 (37)0.5 Mucosal 6 423 (46) 401 (67)1.0 Mucosal 8 476 (52) 189 (33)*5.0 Mucosal 10 398 (35) 160 (27)*1-0 Serosal 9 431 (32) 413 (39)

Glycine 5.0 Mucosal 9 247 (26) 36 (4)*

Pangbourne. Diagnostic kits for the deter-mination of alkaline phosphatase activity wereobtained from Sigma Chemical, Poole andBiorad protein assay reagents from Biorad,Hemel Hempstead. All other reagents were ofanalytical grade.

Expression of resultsResults are expressed as mean (SEM) values ofthe number of observations indicated andStudent's t test, paired or unpaired asappropriate, was used to assess the significanceof peppermint oil action. A kinetic analysis ofthe nutrient dependent rise in SCC wasperformed using the Eadie-Hofstee plot2' tocalculate the apparent transport constant andthe maximum increase in SCC (SCCmax).

Results

Effect ofpeppermint oil on transintestinal electricalactivityControl intestinal sheets generated a basal PDof 2.7 (0. 1) mV and a SCC of 104 (4) [pA/cm2(n= 100), the serosal side being positive withrespect to the mucosal side. Tissue resistancewas 27 (1) ohm.cm2. Values for the test sheetsdid not differ significantly (p>0 05 in allcases). Application of peppermint oil to eitherthe mucosal (up to 5 mg/ml) or serosal solution(1 mgfml) did not significantly affect basalelectrical activity (p>0-05 in all cases).The addition of glucose to the luminal

surface of the intestinal sheets caused a con-centration dependent rise in SCC (Fig 1) thatconsists of two components: an increaseassociated with active sodium dependentglucose absorption and a decrease because ofthe imposition of an osmotic gradient.22 Themagnitude of the osmotic component can beassessed by determining the effect of equimolarmannitol on the SCC (Fig 1), which reducedthe SCC by 24 [iA/cm2/10 mM. The increasein SCC due to active glucose absorption cantherefore be calculated.

Peppermint oil, at concentrations of 1 and5 mg/ml, inhibited the increase in SCCassociated with active glucose absorption,although lower concentrations (0 1 and 0.5mg/ml) were without effect (Fig 1). Pepper-mint oil also reduced the effects of mannitol,with significant differences being observed athigher mannitol concentrations with pepper-mint oil concentrations of 0.5, 1, and 5 mg/ml(Fig 1).A kinetic analysis of the data showed that

mucosal peppermint oil reduced the SCCmax,causing a 59 (5)O/o (n=9) inhibition at 1 mg/mland a 58 (6)% (n= 10) inhibition at 5 mg/ml(Table).Mucosal peppermint oil caused a more pro-

nounced inhibition of the rise in SCC inducedby the actively transported amino acid glycine,reducing the SCCmax by 85 (1)% (n=9) at aconcentration of 5 mg/ml (Table).To find out if the actions of mucosal pepper-

mint oil were selective for nutrient absorption,its effects on acetylcholine induced secretion

216

on May 21, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.39.2.214 on 1 A

ugust 1996. Dow

nloaded from

Influence ofpeppermint oil on absorptive and secretory processes in rat small intestine

' 1250

Cl.

E 100

0

E75

0

25

0o 1 2 3 4 5 6

Inverse osmolality (Osmol-1 Kg H20)Figure 2: Effect of medium osmolality on glucose uptake by brush border membrane vesi£at equilibrium. Each point represents the mean (SEM) offour observations and thecalculated regression line has been extrapolated to infinite osmolality.

1000 _

o 800 -L.Cl.0)

U 600ECL

cc 400

CL)

200

0~

0 1 2 5Time (min)

Figure 3: Time course for glucose uptake into brush border membrane vesicles, with 0(n= 6) representing total uptake and A (n= 6) representing passive glucose uptake in thipresence of 025 mM phlorrhizin. Each point represents the mean (SEM) of the numberof observations indicated.

were investigated. In contrast with its inhibitiof nutrient absorption, peppermint oil in 1

mucosal solution at the highest concentratitested (5 mg/ml) failed to influence the riseSCC induced by serosal 10-3M acetylcholi(control: 249 (27) jiA/cm2, n=9; +peppermoil: 213 (23) puA/cm2, n=9, p>0 05).The effects of the serosal application

peppermint oil were also examined.1 mg/ml peppermint oil had no effect on baelectrical activity, nor did it inhibit the riseSCC associated with active sodium dependeglucose absorption (Table). It did howevreduce the fall in SCC induced by mannitola mean of 31 (2)% (n=9) over the rangemannitol concentrations (p<0.05 at ea

concentration). Serosal peppermint oil a]inhibited the response to 10-3M acetylcholiby 64 (7)% (n= 12) (control: 263 (17) ,uA/cnn= 12; +peppermint oil: 88 (14) pA/cnn=12, p<0 001).

Characterisation of brush border membranevesiclesVesicles showed a 23 (1)-fold (n= 18) enricment in alkaline phosphatase activity and a

(1)-fold (n=8) increase in sucrase activitycompared with the original homogenate.Conversely, Na+/K+-ATPase activity wasreduced in vesicles to 71 (17)% (n=14) ofhomogenate values, while succinate dehydro-genase activity in the vesicle preparation wasreduced by 100 (0)O/o (n=4), demonstrating apurification of the brush border membraneover other cellular membrane types. Non-specific loss of enzyme activity throughout thepreparation was similar for both sucrase (26(5)%, n=4) and Na+/K+-ATPase (36 (4)%,n=4), suggesting that differences in the enrich-ments for these two enzymes are unlikely to be

j caused by selective degradation processes.7 Incubating brush border membrane vesicles

in incubation media of increasing osmolalitycles caused a reciprocal decrease in the amount of

glucose uptake (Fig 2). Extrapolation of thesedata to an infinite osmolality provides anassessment of non-specific binding of glucoseto the brush border membrane (6.9 pmolglucose/mg protein, n=4, r=0.787, p<0 01,Pearson's product-moment correlation coeffi-cient). This value is only 1% of the maximumglucose uptake at 20 seconds (Fig 3), showingthat non-specific binding of glucose makes anegligible contribution to the uptake data.

Effect ofpeppermint oil on glucose uptake bybrush border membrane vesiclesIn the presence of an inwardly directed sodiumgradient glucose was actively taken up by the

, brush border membrane vesicles, a processz that was abolished in the presence of phlor-15 rhizin. Total glucose uptake reached a maxi-

mum at 20 seconds and by 15 minutes it hadfallen to values that did not differ from passive

e uptake observed with phlorrhizin - that is, thereaction had reached equilibrium (Fig 3). Themean peak overshoot for glucose uptake was8.2 (1.6)-fold (n=6) greater than equilibrium

ion values, with intravesicular volume being esti-the mated as 0-83 (0.1) gl/mg protein (n=16).ion Subtraction of passive uptake from the totalin uptake provides a value for net active sodiumine coupled glucose uptake. The presence of pep-mint permint oil in the extravesicular solution

reduced significantly active glucose uptake atof 10 seconds by 45 (10)% (n=6, p<005) atAt 0.5 mg/ml and by 65 (7)% (n=6, p<0 001) atLsal 1 mg/ml. At 005 and 0 1 mg/ml it was withoutin effect (p>0 05 in both cases, Fig 4). Pepper-znt mint oil (0 05 and 1 mg/ml) did not however,er, cause a significant inhibition of glucose uptakeby at the 15 minute time point, indicating that itof had not influenced passive glucose entry intoch the vesicles (Fig 4). This was in contrast withiso the effects of Triton X-100, an agent knownine to disrupt membrane structure,23 whichn2, significantly reduced glucose uptake atn2, equilibrium (Fig 4) to a value (11 (4) pmol

glucose/mg protein, n= 5) that was similar tothat predicted for non-specific binding (Fig 2).

Discussionfh- Peppermint oil is used to treat disturbances of10 intestinal motility. As it is given orally it will

217

on May 21, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.39.2.214 on 1 A

ugust 1996. Dow

nloaded from

Beesley, Hardcastle, Hardcastle, Taylor

800 r A

600 K

400 [-

200 H

oControl

*

**

0.05 0.1 0.5Peppermint oil concentration (mg/mi)

100r B

75K

50

25

Control 0.05

T*

1.0 Triton X-10Peppermint oil concentration (mg/mi)

Figure 4: Effects ofpeppermint oil or Triton X-100 on (A) active sodium dependeuptake into brush border membrane vesicles at 10 seconds and (B) total glucose u,equilibrium (15 minutes). Each point represents the mean (SEM) for n= 6 (contipeppermint oil) or n=5 (Triton X-100) observations and a paired t test was usedthe significance ofpeppermint oil or Triton X-100 action. *p<0.05, **p<0.0O1.

come into contact with the intestinalwhere it may influence the transport acthe enterocytes. This study shows thatmint oil can affect both the absorptsecretory functions of the intestine,which seem to be specific to the tiprocesses investigated and not the r

non-selective tissue damage.Peppermint oil, when applied to ei

mucosal or serosal side of an intestinpreparation, had no effect on basal e

activity, suggesting that basal electrog4transport systems were unaffected. It dever, reduce the electrokinetic pinduced by the imposition of an osmotient across the tissue, an effect that wwith both mucosal and serosal pepperapplication. Adding an impermeant scthis case mannitol, to the mucosal %

leads to a decrease in the PD22 acconby a fall in the SCC. This arises from 1

flow of fluid across the tissue, which dmobile ions through charged juinduces a boundary diffusion potentiaunstirred layer immediately adjacentmembrane, and generates a profile asypotential by distorting ion conce.profiles within the membrane.24 Theelectrokinetic effect seen in the pres

either mucosal or serosal peppermint oilcannot be attributed to a non-specific increasein the permeability of the intestinal sheet as

00% tissue resistance remained unchanged. As the.o effects on the electrokinetic potential were seenQ when peppermint oil was present on either thes mucosal or serosal side of the tissue it is

, probable that it alters the charge on the tight45% junctions, which are considered to be the route

a by which bulk flow of fluid across the tissue_65% a) occurs.25

' Mucosal, but not serosal, peppermint oilinhibited the rise in SCC obtained on additionof glucose. This, when corrected for the

100% osmotic component, has been shown to reflectthe rate of active sodium linked glucoseabsorption26 and this study indicates thatpeppermint oil inhibits this process. Similareffects were seen with the actively transportedamino acid glycine. A kinetic analysis of thedata shows that peppermint oil causes a signifi-cant fall in the SCCmax, suggesting that thecapacity of the transport system had beenimpaired. The effects of peppermint oil on theapparent transport constant were inconsistent.There are however, several difficulties in theinterpretation of kinetic data derived frommeasurements of intestinal transport.24 Firstlythere is a significant unstirred layer at theluminal membrane that impedes access of thenutrient to its carrier sites, leading to artificiallyhigh values for the apparent transportconstant. Moreover, changes in SCCmax can

DO also lead to an alteration in the calculated valuefor the apparent transport constant.The ability of peppermint oil to inhibit

'nt glucose nutrient absorption seems to represent a direct

pol and action on the membrane transport process as itto assess was also observed in brush border membrane

vesicles where the intracellular machinery isabsent. Peppermint oil caused,a concentration

mucosa dependent reduction in sodium dependent2tivity of glucose uptake, the maximum level of inhibi-pepper- tion seen being similar to that observed in thetive and Ussing chamber experiments, suggesting aeffects common mode of action. It does not seem to

ransport act by disrupting membrane integrity as the-esult of passive uptake of glucose into brush border

membrane vesicles was unaltered, in contrastither the with the effects of Triton X-100, which istal sheet known to adversely affect membrane structure..lectrical It may, therefore, have a more specific effect on,enic ion the active transport mechanism. This actionlid how- may be exerted at the sodium site of the?otential glucose carrier SGLT1,27 as glycine absorp-ic gradi- tion, which is also a sodium dependentvas seen process, was similarly inhibited by peppermintmint oil oil. Alternatively, any incorporation of theDlute, in lipophilic peppermint oil into the luminalsolution membrane could influence the rate at whichripanied transporters function.the bulk Although a detailed investigation into thelisplaces actions of peppermint oil on intestinal secre-nctions, tory processes was not undertaken, the factal in the that secretion induced by acetylcholine was notto the inhibited by mucosal peppermint oil indicates

rmmetry that the effects of peppermint oil on brushntration border membrane nutrient transport werereduced selective. The secretory process entails thegence of opening of chloride channels in the luminal

c

._

0)0a

E

0.E

a)Q0.

a)CO0c.)

c

0.a

0)0Ei-.0E

a)0

a)Co00

(:3U

218

on May 21, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.39.2.214 on 1 A

ugust 1996. Dow

nloaded from

Influence ofpeppermint oil on absorptive and secretory processes in rat small intestine 219

membrane, allowing chloride ions accumu-lated within the enterocyte to diffuse into thelumen down their electrochemical gradient.28It has been argued that chloride secretion is afunction of crypt cells, while absorption occursin villous cells.29 More recent evidence how-ever, indicates that villous cells also possess thecapacity for secretion. A secretagogue induceddepolarisation of the luminal membrane hasbeen observed in villous cells30 and in situhybridisation has shown that M3 cholinorecep-tors, the subtype responsible for cholinergicallyinduced secretion, are localised to the villi.31The fact that mucosal peppermint oil has noaction on secretion, even though chloridesecretion and nutrient absorption may becolocalised, is an indication of the selectivity ofits actions.

Serosal peppermint oil had no effect onactive sodium linked glucose absorption as therise in SCC associated with this process wasunaffected. In contrast, the presence ofpeppermint oil on this side of the tissue didcause a considerable inhibition of acetylcholineinduced secretion. Calcium is important forintestinal secretion,13 14 and the fact that theacetylcholine induced response is reduced bythe removal of serosal calcium or the presenceof the calcium channel blocker verapamil inthe serosal solution32 suggests that thissecretagogue acts by increasing calcium entryinto the cell via calcium channels at the baso-lateral membrane. Peppermint oil has beenshown to limit the availability of calcium,6 aneffect that can be explained by its actions onpotential dependent calcium channels.7 Thefact that peppermint oil can inhibit specific[3H] nitrendipine and [3H] PN 200-1 10 bind-ing to smooth muscle and neuronal prepara-tions suggest that it can cause a selective blockof these channels.6 The ability of peppermintoil to block these channels would explain itsinhibition of the secretory response.

This study shows that peppermint oil iscapable of influencing the transport activity ofthe enterocytes that line the intestinal lumen.The standard bolus dose of peppermint oil forhuman subjects is about 400 mg. In the fastingstate a local concentration of 4 mgfml in theintestinal lumen could readily be achieved, avalue that was effective in causing a change inintestinal transport. The fact that constituentsofpeppermint oil appear in the urine8 indicatesthat it must eventually cross the intestinalepithelium and would therefore be capable ofexerting actions on secretory mechanisms aswell as influencing absorption when it is pre-sent at the luminal surface.The authors would like to thank both Dr E Debnam and DrS P Shirazi-Beechey for their invaluable assistance with thebrush border membrane vesicle work. AB was funded by theSpecial Trustees for the Former United Sheffield Hospitals.

1 Rees WDW, Evans BK, Rhodes J. Treating irritable bowelsyndrome with peppermint oil. BMJ 1979; 2: 835-6.

2 Taylor BA, Luscombe DK, Duthie HL. Inhibitory effect ofpeppermint oil on gastrointestinal smooth muscle. Gut1983; 24: A992.

3 Taylor BA, Duthie HL, Luscombe DK. Mechanism bywhich peppermint oil exerts its relaxant effect on gastro-intestinal smooth muscle. J Pharm Pharmacol 1985; 37:104P.

4 Hills JM, Potts HJ, Carlin BA, Parsons ME. An examina-tion of the contribution made by the constituents ofpeppermint oil (PO) to the relaxation of gastrointestinalsmooth muscle. BrJPharmacol 1990; 101: 600P.

5 Leicester RJ, Hunt RH. Peppermint oil reduces colonicspasm during endoscopy. Lancet 1982; ii: 989.

6 Hawthorn M, Ferrante M, Luchowski E, Rutledge A, WeiXY, Triggle DJ. The actions of menthol and peppermintoil on calcium channel dependent processes on intestinal,neuronal and cardiac preparations. Aliment PharmacolTher 1988; 2: 101-18.

7 Hills JM, Aaronson PI. The mechanism of action ofpepper-mint oil on gastrointestinal smooth muscle.Gastroenterology 1991; 101: 55-65.

8 Somerville KW, Richmond CR, Bell GD. Delayed releasepeppermint oil capsules (Colpermin) for the spastic colonsyndrome: a pharmacokinetic study. Br J Clin Pharmacol1984; 18: 638-40.

9 Fondacaro JD. Intestinal ion transport and diarrhealdisease. AmJ'Physiol 1986; 250: G1-8.

10 Lundgren 0. Nervous control of intestinal transport.Baiiere's Clin Gastroenterol 1988; 2: 85-106.

11 Field M, Rao MC, Chang EB. Intestinal electrolyte trans-port and diarrheal disease. N Engl J Med 1989; 321:800-6.

12 Perdue MH, McKay DM. Integrative immunophysiology inthe intestinal mucosa. Am Jr Physiol 1994; 267: G151-65.

13 Donowitz M. Ca2+ in the control of active intestinal Na andCl transport: involvement in neurohumoral action. Am JPhysiol 1983; 245: G165-77.

14 Donowitz M, Welsh MJ. Ca2+ and cyclic AMP in regula-tion of intestinal Na, K, and Cl transport. Ann Rev Physiol1986; 48: 135-50.

15 Beesley A, Hardcastle J, Hardcastle PT, Taylor CJ.Peppermint oil (PO) reduces Na+-dependent glucoseabsorption in rat small intestine. Gut 1994; 35: T217.

16 Field M, Fromm D, McColl I. Ion transport in rabbit ilealmucosa. I. Na and Cl fluxes and short-circuit current.AmJPhysiol 1971; 220: 1388-96.

17 Kessler M, Acuto 0, Storelli C, Murer H, Muller M,Semenza G. A modified procedure for the rapid prepara-tion of efficiently transporting vesicles from small intesti-nal brush border membranes. Biochim Biophys Acta 1978;506: 136-54.

18 Dahlqvist A. Assay of intestinal disaccharidases. AnalytBiochem 1968; 22: 99-107.

19 Hardcastle J, Hardcastle PT, Cookson J. The inhibitoryactions of loperamide on absorptive processes in rat smallintestine. Gut 1986; 27: 686-94.

20 Pennington RJ. Biochemistry of dystrophic muscle.Mitochondrial succinate-tetrazolium reductase andadenosine triphosphatase. Biochem J 1961; 80: 649-54.

21 Engle PC. Enzyme kinetics. 2nd ed. London: Chapman andHall, 1981: 14-25.

22 Smyth DH, Wright EM. Streaming potentials in the ratsmall intestine. J Physiol 1966; 182: 591-602.

23 Collins MLP, Salton MRJ. Solubility characteristics ofMicrococcus lysodeikticus membrane components indetergents and chaotropic salts analyzed by immunoelec-trophoresis. Biochim Biophys Acta 1979; 553: 40-53.

24 Levin RJ. Fundamental concepts of structure and functionof the intestinal epithelium. In: Duthie HL, WormsleyKG, eds. Scientific basis of gastroenterology. Edinburgh:Churchill Livingstone, 1979: 308-37.

25 Diamond JM. Channels in epithelial cell membranes andjunctions. Fed Proc 1978; 37: 2639-44.

26 Luppa D, Hartenstein H, Muller F. Relation betweenmicrovilli membrane potential and glucose transportcapacity of rat small intestine. Biomed Biochim Acta 1987;46: 341-8.

27 Wright EM. The intestinal Na+/glucose cotransporter.Ann Rev Physiol 1993; 55: 575-89.

28 Hardcastle J, Hardcastle PT. Membrane permeabilitychanges in intestinal transport. Med Sci Res 1987; 15:471-3.

29 Ewe K. Intestinal transport in constipation and diarrhoea.Pharmacology 1988; 36 (suppl 1): 73-84.

30 Stewart CP, Turnberg LA. A microelectrode study ofresponses to secretagogues by epithelial cells on villus andcrypt of rat small intestine. Am JT Physiol 1989; 257:G334-43.

31 Przyborski SA, Levin RJ. Enterocytes on rat jejunal villibut not in the crypts possess m3 mRNA for the M3muscarinic receptor localized by in situ hybridization.Exp Physiol 1993; 78: 109-12.

32 Hardcastle J, Hardcastle PT, Noble JM. The involvement ofcalcium in the intestinal response to secretagogues in therat.JTPhysiol 1984; 355: 465-78.

on May 21, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.39.2.214 on 1 A

ugust 1996. Dow

nloaded from